Thank you very much!  This helps a lot.  I think I am ready to render now.

Jon


On Wed, 3 Dec 2003, Matto Marjanovic wrote:

>
>  >I think I've figured it out, but if someone could verify my results, that
>  >would be great.  I think I need to design at 1365x768.  Using the
>  >information from the VCD stills page and other pages I found, I built a
>  >little aspect ratio converter that, given the number of pixels high and
>  >wide the destination platform is, and the display aspect ratio, it will
>  >calculate what width/height you should design at.
>  >
>  >If anyone wants to offer comments, that would be great, as dealing with
>  >the various aspect ratios has me tipping on the edge of insanity.
>
> Hiya,
>
> My first instinct is to refer you to http://www.mir.com/DMG/aspect.html
>  (which I myself am now turning to...).
>
>     width      DAR         1024   16/9             16 * 768    4
>    -------- = -----  -->   ---- = ----- --> SAR = ---------- = -
>     height     SAR          768    SAR             9 * 1024    3
>
> So, you're target material has a sample aspect ratio of 4:3 (which
>  evokes a big retrospective "Duh", because you are taking a standard
>  4:3 screen and stretching the pixels by 4:3 to create a 16:9 screen).
>
>
> The options that I can think of:
>
>  1) Use y4mscaler to do the scaling.  This should do it:
>
>      ... | y4mscaler -O size=1024x768 -O sar=4:3 | ...
>
>     You can start with any input and it should scale it to look correct
>     on that funky screen, but "1365.3-x768" is an ideal source size (with
>     a 1:1 SAR) if you want to fill the whole screen.
>
>  2) You are rendering an animation, right?  If it is a decent renderer,
>     you should be able to specify both the frame size and the sample aspect
>     ratio.  This is ideal, because the machine only renders the pixels that
>     you actually need and you can avoid any post-process scaling.
>
>
> If the screen/stream is going to be interlaced, you will want to avoid any
>  vertical scaling.  If the renderer is classy enough to synthesize an
>  interlaced output, then (2) should be a viable option, anyhow.  (Although,
>  if this is going through Windows Media Player, I guess interlacing is
>  probably not anywhere in the equation.)
>
>
> Another thought --- are you saying that the Windows running through this
>  big wide screen looks widened?  I would think that Windows sees the screen
>  as 1365x768, outputs a 1:1 SAR frame of that size, and a scaler built into
>  display actually smooshes the pixels together to put it on the glass.
>  Or does Windows actually produce a non-1:1 SVGA output?  I would think
>  "No way" --- and every display on the market is going to assume that its
>  *input* is 1:1 pixels, just because nobody understands non-square pixels
>  anyway.
>
>  If all this is the case... just render at 1365x768 with 1:1 pixels and
>  you are done.  (Phew!)
>
> -matt m.
>



-------------------------------------------------------
This SF.net email is sponsored by OSDN's Audience Survey.
Help shape OSDN's sites and tell us what you think. Take this
five minute survey and you could win a $250 Gift Certificate.
http://www.wrgsurveys.com/2003/osdntech03.php?site=8
_______________________________________________
Mjpeg-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/mjpeg-users

Reply via email to